Cryptographic Security Concerns in Quantum Key Distribution Systems
Quantum key distribution protocols are known to be vulnerable against some side channel attacks. The photon detectors are known to be the “Achilles' heel” in security of the quantum key distribution (QKD). We introduce two new methods for attacking the QKD systems in this study. The first attack involves the detection of RF radiation from the avalanche photo-diodes (APDs). The radiation can be detected via an ultra-wide band antenna and the grouping of the signals depending on the detector is done via a machine learning algorithm. While supervised learning is used for this study, it is possible to extend the work to unsupervised learning. The second attack we study is about creating an extra delay in one of the detectors’ timestamps by externally heating it up by injecting noise with a known polarization. The heating up of the detector causes a shift in the cross-correlogram, which can be identified by third parties. We also present the prevention methods for both these two attacks.